wait-what

Pass

Audited by Gen Agent Trust Hub on Aug 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains no executable scripts, binaries, or code files. All instructions are provided in natural language within the markdown body, which inherently limits the attack surface for categories such as remote code execution or persistence.
  • [INDIRECT_PROMPT_INJECTION]: The skill directs the agent to utilize language and context from an external file, CONTEXT.md. This creates a surface for indirect prompt injection if that file contains untrusted content, though it is standard functionality for this skill's purpose.
  • Ingestion points: Instructions in SKILL.md reference CONTEXT.md for context and ubiquitous language.
  • Boundary markers: None present.
  • Capability inventory: The skill explicitly sets disable-model-invocation: true, which prevents the agent from calling external tools or executing commands while this skill is active, providing a strong security boundary.
  • Sanitization: None present.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 19, 2026, 05:19 PM
Security Audit — agent-trust-hub — wait-what